Objective: Legal decisions in sexual assault cases often hinge on the presence or absence of genitorectal injury. Unfortunately, the forensic literature does not explain why some victims sustain genitorectal injury and others do not. This study explores possible predictors of genitorectal injury in adult female sexual assault victims. Methods: This retrospective cross-sectional analysis forms the derivation set for a larger planned prospective analysis. The authors extracted data describing consecutive female sexual assault victims who met inclusion criteria between July 1995 and July 1998. Exclusion criteria included male sex, lack of estrogen in females, consensual intercourse within the previous 72 hours, and lack of penetration during the assault. The authors explored associations between genitorectal injury and seven demographic variables, nine assault characteristics, and the time between assault and exam or postcoital interval (PCI). Variables thought to be predictive were incorporated into a logistic regression model. Results: Five hundred forty-eight sexual assault victims were seen during the study time period; 209 of these met the inclusion criteria. Logistic regression controlling for important covariates showed an increase risk of genitorectal injury with a PCI < 24 hours (OR 7.47, 95% CI = 1.78 to 31.35), physical/verbal resistance (OR 5.96, 95% CI = 1.21 to 29.36), rectal penetration (OR 7.47, 95% CI = 1.05 to 53.07), and greater than high school education (OR 7.13, 95% CI = 1.03 to 49.65). Conclusions: This study presents an important first look at variables that may predict genitorectal injury in sexual assault victims. Future studies that examine more data are needed to corroborate this preliminary derivation set analysis.